In this proposal we seek to develop a new method and instrument to reduce the complexity of protein mixtures derived from human fluids with an eye to discovery of very low-abundance protein biomarkers. The proposed method accomplishes two tasks: it reduces the dynamic range of the protein mixture by three or more orders of magnitude and it concentrates proteins by a factor of 100 or more so that low- abundance can be detected and concentration differences noted. The method does not use antibodies, combinatorial peptide libraries, carrier ampholytes or solid-liquid adsorption equilibria such as HPLC. The samples produced are compatible with 2D gel electrophoresis and 2D liquid chromatography tandem mass spectrometry approaches. By using different physical methods to separate the protein mixtures, the goal is to uncover new biomarkers that have eluded existing state-of-the-art methods. After development, the proposed method and instrument will be demonstrated using human plasma, where a sample from a lung cancer patient is compared with a reference that does not have cancer. Additional low-abundance proteins that are potential biomarkers will be uncovered when compared with traditional immunodepletion methods.